Brand oriented list management engine in item list system

Abstract

Methods, systems, and computer storage media for providing brand oriented list management in an item list system are described. By using a brand-oriented machine learning model, the brand-oriented list management engine supports list accuracy in compliance with item list market criteria and optimizes search visibility and customer experience. The brand-oriented list management engine also supports identifying product list guides that intentionally bypass or violate a project list system, and taking corrective actions to improve list quality and functionality associated with list management, including seller list flows and optimizing visibility based on list quality signals. In operation, an item list associated with an item list system is accessed. The list is analyzed using a brand-oriented machine learning model trained based on a multi-dimensional authenticity analysis dataset. Based on analyzing the list of items, brand-oriented security notifications associated with the list of items are generated. The brand-oriented security notification is transmitted.

Description

Technical Field

[0001] The various aspects of the technology described in this article generally relate to systems, methods, and computer storage media for providing brand-oriented AI management, etc., using brand-oriented AI systems associated with project listing systems. Background Technology

[0002] Users utilize listing systems to browse product listings, make purchases, and sometimes even sell items. When sellers list items (especially branded items), they must adhere to guidelines set by the platform and ensure the authenticity and accuracy of their listings. For users buying items, branded items typically enjoy a reputation for quality and authenticity, thus influencing their purchasing decisions. However, listing counterfeit or misleading items damages the trust between buyers and the listing system. In this context, listing providers implement strategies to prevent fraudulent listings, protect brand integrity, and maintain a positive user experience to safeguard their reputation and legal obligations. For example, brand security tools can include trademark monitoring, anti-counterfeiting measures, digital brand protection, intellectual property enforcement, and market surveillance. Summary of the Invention

[0003] The brand-oriented AI system provides brand-oriented AI capabilities to support various applications and services within the project listing system. This system includes a brand-oriented machine learning model training engine, a brand-oriented security administrator engine, and a brand-oriented listing management engine. In this way, the brand-oriented AI system supports the project listing system by automating brand-related tasks such as machine learning training, security management and analysis, and project listing quality execution.

[0004] The brand-oriented machine learning model training engine supports training brand-oriented machine learning models that predict brands in item listings that do not include brand information. Model training includes brand data processing, model selection, and optimization algorithms. Specifically, brand-oriented machine learning model training is the process of teaching a machine learning model (using a multidimensional authenticity analysis dataset) to identify patterns and make predictions or decisions based on input data. This training can be based on novel training techniques and training features in data associated with multidimensional authenticity analysis datasets (i.e., brand protection and verification data and item listing system data) to generate brand-oriented machine learning models. For example, the training is based on brand protection and verification data (e.g., verified rights holder (VeRO) data and after-sales certification (PSA) data) and item listing system data (e.g., item listing text and / or images) that support the identification of highly relevant brand features for machine learning to support brand prediction for item listings.

[0005] Brand-oriented machine learning models can be multimodal models, including text-based models and image recognition models, which can be selectively implemented to support functionality (e.g., applications in a list system). Brand-oriented machine learning models can be deployed to operate in conjunction with applications and services associated with a brand-oriented security administrator engine and a brand-oriented list management engine.

[0006] The Brand-Oriented Security Administrator Engine supports the generation of comprehensive reports and analyses on brand-related security metrics, including brand compliance rates, infringement incidents, enforcement actions taken, and overall project list system integrity. The Brand-Oriented Security Administrator Engine also supports the visualization of brand-oriented security analytics results data by presenting complex brand-oriented security analytics results data in a clear, actionable format, enhancing usability and decision-making.

[0007] The brand-oriented product listing management engine ensures listing accuracy meets market standards and optimizes search visibility and customer experience. It also identifies products that intentionally circumvent or violate the product listing system's guidelines and takes corrective action to improve listing quality. Furthermore, the engine enhances customer satisfaction by improving CSAT scores; minimizes potential losses and attracts more customers to the product listing platform by reducing negative buying experiences; and promotes growth in the total transaction value within the product listing system platform.

[0008] In this way, the brand-oriented machine learning model training engine trains and deploys brand-oriented machine learning models; the brand-oriented security administrator tool utilizes novel brand-oriented machine learning models to provide novel functionalities associated with brand monitoring and detection, dashboards, real-time alerts and notifications, and trend analysis graphs; and the brand-oriented listing management tool utilizes novel brand-oriented machine learning models to provide novel functionalities associated with listing management, including seller listing processes and optimizing visibility based on listing quality signals.

[0009] This summary is provided to describe in simplified form the selection of concepts further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to help define the scope of the claimed subject matter. Attached Figure Description

[0010] The technology described herein will now be described in detail with reference to the accompanying drawings, in which:

[0011] Figure 1A and Figure 1BThis is a block diagram of a brand-oriented AI system for providing brand-oriented AI system management in a project listing system, based on various aspects of the technology described in this article.

[0012] Figures 2A-2C This is a diagram and interface illustrating the various aspects of the technology described in this article and their connection to providing brand-oriented AI system management in a project listing system;

[0013] Figure 3 It is an interface related to the various aspects of the technology described in this article and the provision of brand-oriented artificial intelligence system management in the project listing system;

[0014] Figure 4 It is an interface related to the various aspects of the technology described in this article and the provision of brand-oriented artificial intelligence system management in the project listing system;

[0015] Figure 5 It is an interface related to the various aspects of the technology described in this article and the provision of brand-oriented artificial intelligence system management in the project listing system;

[0016] Figure 6 It is an interface related to the various aspects of the technology described in this article and the provision of brand-oriented artificial intelligence system management in the project listing system;

[0017] Figures 7A-7C It is an interface related to the various aspects of the technology described in this article and the provision of brand-oriented artificial intelligence system management in the project listing system;

[0018] Figure 8 It is an interface related to the various aspects of the technology described in this article and the provision of brand-oriented artificial intelligence system management in the project listing system;

[0019] Figures 9A-9B It is an interface related to the various aspects of the technology described in this article and the provision of brand-oriented artificial intelligence system management in the project listing system;

[0020] Figure 10 This paper provides a first exemplary method for providing brand-oriented machine learning model training engine functionality in a project listing system based on various aspects of the techniques described herein;

[0021] Figure 11 A second exemplary method is provided for providing brand-oriented machine learning model training engine functionality in a project listing system based on various aspects of the techniques described herein;

[0022] Figure 12A third exemplary method is provided for providing brand-oriented machine learning model training engine functionality in a project listing system based on various aspects of the techniques described herein;

[0023] Figure 13 This paper provides a first exemplary method for providing brand-oriented security administrator engine functionality in a project listing system based on various aspects of the technology described herein;

[0024] Figure 14 A second exemplary method is provided for providing brand-oriented security administrator engine functionality in a project listing system based on various aspects of the technology described herein;

[0025] Figure 15 A third exemplary method is provided for providing brand-oriented security administrator engine functionality in a project listing system based on various aspects of the technology described herein;

[0026] Figure 16 This paper provides a first exemplary method for providing brand-oriented list management engine functionality in a project list system based on various aspects of the technology described herein;

[0027] Figure 17 A second exemplary method is provided for providing brand-oriented list management engine functionality in a project list system based on various aspects of the technology described herein;

[0028] Figure 18 A third exemplary method is provided for providing brand-oriented list management engine functionality in a project list system based on various aspects of the technology described herein;

[0029] Figure 19 A list of exemplary projects and a block diagram of a system computing environment suitable for implementing various aspects of the techniques described herein are provided;

[0030] Figure 20 Block diagrams are provided of exemplary distributed computing environments suitable for implementing various aspects of the techniques described herein; and

[0031] Figure 21 This is a block diagram of an exemplary computing environment suitable for implementing various aspects of the techniques described herein. Detailed Implementation

[0032] Overview

[0033] The project listing system and platform support storing projects (products or assets) in a project database and provide a search system for receiving queries and identifying search result projects based on those queries. A project (e.g., a physical project or a digital project) refers to a product or asset offered for listing on the project listing platform. The search system supports identifying result projects from the project database in response to received queries. The project database can be dedicated to a content platform or project listing platform, such as the eBay content platform developed by eBay Inc. in San Jose, California. The project listing system can also provide AI-enabled applications (“AI applications”) that utilize AI models (e.g., natural language processing (NLP) models and image recognition models) to perform computational tasks. Applications use AI to perform a variety of tasks across different domains, including product recommendation, search optimization, sentiment analysis, product categorization, customer support, and spam and fraud detection.

[0034] In the fast-paced world of e-commerce, ensuring product listings are complete and accurate is crucial for sellers, buyers, and suppliers on e-commerce platforms. For example, experienced online seller Jane is preparing to list a new set of wireless headphones on her preferred listings platform. However, in her haste, she forgets to fill in the "Brand" field before submitting the listing. This seemingly insignificant oversight can lead to a range of technical and user experience issues that can impact everything from search functionality to customer trust. Omitting brand information from item listings on listing platforms presents significant technical challenges in combating counterfeiting. Automated detection systems struggle to verify product authenticity without brand data, while manual review is more labor-intensive and error-prone. Incomplete brand information hinders trend analysis and reporting, creating data gaps that impact insights into counterfeit activity. This data deficiency damages user trust, complicates compliance with laws and regulations, and reduces the effectiveness of user reporting mechanisms. To address these issues, platforms must enforce brand fields, enhance verification processes, educate sellers, and work closely with brands.

[0035] Listing system providers implement strategies to prevent fraudulent listings, protect brand integrity, and maintain a positive user experience to uphold their reputation and legal obligations. Brand security can refer to measures taken within online marketplaces to protect a brand's identity, reputation, and intellectual property. It involves preventing various threats, including counterfeiting, unauthorized resale, trademark infringement, brand impersonation, and brand dilution. Brand security can include trademark protection, counterfeiting prevention, authorized seller management, brand monitoring, brand identity protection, cybersecurity, and legal enforcement.

[0036] Typically, project listing systems lack the comprehensive logic and infrastructure to effectively provide brand-oriented AI capabilities. These systems fail to adequately leverage AI for brand-oriented security machine learning model training, security administrator tools, and listing management tools.

[0037] The project listing system lacks the machine learning engine, technology, and data to adequately train machine learning models to effectively and comprehensively support the system. Machine learning model training may be based on limited and biased data, thus preventing the model from fully generalizing to a wide variety of brands. Furthermore, without sufficient contextual understanding (e.g., understanding brands based on industry, market trends, and seller and buyer behavior), the trained model may misinterpret brand-related signals or fail to capture subtle brand associations.

[0038] The item listing system also lacks adequately trained machine learning models to support brand-oriented security. For example, brand security refers to measures taken within the item listing system to protect a brand's identity, reputation, and intellectual property. The item listing system can collect data based on item categories and other data focusing on brand characteristics. However, without appropriate brand security data and brand monitoring tools, security administrators cannot visualize and identify fraud trends to safeguard the brands associated with the item listing system.

[0039] Listing systems often lack well-trained machine learning models to effectively support listing functionality and features. For example, listing platforms may face challenges in balancing a seamless user experience with accurate listing information. On one hand, they may prefer to require users to enter all necessary data during the listing process, while on the other hand, they may want to streamline the seller listing process without requiring all listing information to be entered. Furthermore, the lack of brand-oriented machine learning models may limit the ability to identify fraudulent listings and reduce their impact on the listing system.

[0040] Therefore, conventional list systems can be improved by leveraging advanced machine learning models and technologies, along with list system tools, to overcome these limitations. These advanced machine learning models and technologies, along with list system tools, can enhance the machine learning capabilities used to protect brands within the list system. Thus, a more comprehensive list system (with an alternative foundation for performing brand-oriented AI management) can improve computational operations and interfaces to provide brand protection within the list system.

[0041] Embodiments of the present invention relate to systems, methods, and computer storage media for providing brand-oriented AI management, etc., using a brand-oriented AI system associated with a project listing system. The brand-oriented AI system provides brand-oriented AI functionality to support various applications and services within the project listing system. The brand-oriented AI system includes a brand-oriented machine learning model training engine, a brand-oriented security administrator engine, and a brand-oriented list management engine.

[0042] Description of technical solutions

[0043] At a high level, listing system providers enforce brand safety policies within their platforms. To streamline the listing process, listing systems may allow sellers to leave certain fields, including the brand field, blank. This omission can occur intentionally or unintentionally. When the brand field is left blank, the system must employ various techniques, such as natural language processing, image recognition, or database cross-referencing, to infer the brand of the item.

[0044] The lack of a designated brand can severely hinder a provider's ability to enforce brand security protocols. For example, implementing brand-specific security measures becomes challenging without a clear brand identity. Furthermore, malicious sellers might intentionally omit brand information to circumvent security policies designed to protect specific brands. This vulnerability necessitates robust brand inference and verification mechanisms to maintain the integrity and security of the item listing system.

[0045] Brand-oriented AI systems provide a machine learning engine for training machine learning models that can predict brands from a list of items excluding brand information. A machine learning training engine is a software system or framework designed to facilitate the process of training machine learning models. It typically includes functions for data preprocessing, model selection, hyperparameter tuning, and algorithm optimization to iteratively improve the model's performance based on a given dataset.

[0046] The machine learning model is trained on various data sources (e.g., item listing system data; verified rights owner (VeRO) data; and after-sales certification (PSA) data) and further trained to predict the brand of an item listing based on item listing information (i.e., indicators) (e.g., title, category, listing site, and image). The machine learning model is trained using algorithms (e.g., word embedding algorithms and image recognition algorithms) to generate a text- and / or image-based machine learning model that supports brand prediction. Importantly, the machine learning is trained on a multidimensional authenticity analysis dataset containing specific training data (i.e., verification and certification data, and historical item listing system data) used to train the brand-oriented machine learning model. In this way, the predicted brand can be identified based on an item listing with missing brand information as input.

[0047] Predicting missing brand information can significantly improve the ability to measure and combat counterfeiting rates, as well as improve the accuracy of counterfeit detection (capture) and false negatives (missed detection). Enhanced counterfeit detection can be based on improved algorithmic accuracy. Predicting and filling in missing brand information improves the quality of the data used to train machine learning algorithms, thus enabling more accurate counterfeit detection. Having brand information ensures consistent data patterns, making it easier for algorithms to identify anomalies that may indicate counterfeit goods.

[0048] Because predicted brand information allows for cross-referencing of the list with existing databases of genuine and known counterfeit products, better cross-referencing is possible, enhancing the identification process. With brand data, the system can more effectively filter and flag suspicious lists, increasing the chances of catching counterfeit products. The counterfeit rate can be measured based on an accurate measure of the brand (e.g., brand A or brand B), not just the category (e.g., handbags or shoes). Predicting missing brand information creates a more complete dataset, providing a more accurate measure of counterfeit prevalence. Improved data allows for more reliable trend analysis, helping to identify patterns in counterfeit activity over time. Comparative analysis can also be performed by comparing counterfeit detection rates before and after brand predictions, and the platform can measure improvements in counterfeit identification. With accurate brand data, the platform can generate brand-specific counterfeit rates, providing valuable insights for both platform administrators and brand owners.

[0049] Furthermore, improvements can be made in identifying captures and omissions because by enabling the algorithm to capture more instances of counterfeit goods, predicting missing brand information reduces the number of reports from third-party or external entities (omissions). Improved data quality gives detection results greater credibility, ensuring fewer counterfeit items go undetected. Better performance tracking is possible because item listing platforms can better track the performance of their detection systems by comparing the rate of counterfeit capture (capture) with the rate of missed counterfeit goods and refining their metrics using predicted brand information. Detailed analysis of captures and omissions allows for continuous improvement of the detection algorithm, making it more effective over time.

[0050] Furthermore, predicting missing brand information significantly improves the quality of item listings in several ways. First, it enhances searchability and discoverability by allowing products to appear in brand-specific searches and filters, thereby improving the overall search experience and potentially increasing sales. Second, complete listings with brand information build user trust and confidence, leading to higher customer satisfaction and fewer returns or disputes. Third, including brand information ensures consistency across all listings, making the platform appear more professional and user-friendly. Fourth, accurate brand information improves recommendation algorithms, thereby enhancing the personalization of the shopping experience and increasing user engagement.

[0051] Furthermore, complete brand information enables the generation of more accurate analytics and insights, providing valuable data for both the platform and sellers to refine their strategies. Additionally, predictive brand information helps detect counterfeit items, reducing the risk of fraudulent listings and enhancing the platform's reputation for hosting genuine products. Finally, accurate brand information ensures compliance with legal and regulatory requirements, helping the platform avoid legal issues and maintain its integrity. Overall, by predicting and filling in missing brand information, the listing system can significantly improve the quality of listings, resulting in a better user experience, increased buyer trust, and improved platform performance and reputation.

[0052] For example, a brand-oriented machine learning model can include one or more machine learning models. A brand-oriented machine learning model can be a word embedding model or a text classification model. Specific word embedding machine learning algorithms can be used as specific implementations within a broader category of word embedding and text classification models (including other models such as FastText, Word2Vec, and GloVe). Machine learning algorithms can be tools for natural language processing tasks, particularly when dealing with large vocabularies, morphologically rich languages, or when computational efficiency is a concern. Word embedding and text classification models are designed to transform words and text into numerical vectors that capture semantic meaning, thereby facilitating a wide range of natural language processing tasks. For example, a machine learning algorithm could be a Word2Vec model, which further includes the ability to process sub-word information, particularly useful for morphologically rich languages ​​and handling words outside the vocabulary.

[0053] Brand-oriented machine learning models can be, or include, image recognition models. Image recognition models can be based on deep convolutional neural networks (CNNs), which are designed to facilitate the training of very deep networks by solving the vanishing gradient problem. For example, ResNet is a deep convolutional neural network (CNN) that excels in image classification and object detection tasks. It introduces the concept of residual learning through shortcut connections, which allow gradients to flow more easily through network layers, making it possible to efficiently train networks with hundreds or even thousands of layers. The ResNet architecture typically consists of multiple residual blocks, where each block contains several convolutional layers and shortcut connections that bypass one or more of these layers, thus directly adding the input to the output. This approach helps maintain performance and accuracy even as network depth increases significantly.

[0054] By using image recognition models and word embedding models, brand-oriented machine learning models (as multimodal models) can process and analyze both textual and visual data, thereby improving their ability to automatically identify brands. Brand-oriented machine learning models can combine text classification capabilities with image recognition capabilities to form multimodal or multi-input models. This approach allows the model to simultaneously process and extract features from both textual descriptions and associated images. Brand-oriented machine learning models can analyze product descriptions and images to provide a comprehensive understanding of the listed items. For example, the text classification part of the model might use natural language processing (NLP) techniques to classify products or analyze customer sentiment, while the image recognition component might employ computer vision techniques to perform tasks such as object detection or brand logo recognition. By integrating both text processing and image processing capabilities, the model can leverage the unique information provided by each modality, thereby improving its ability to make accurate predictions or classifications. This fusion of modalities enables more nuanced insights, such as identifying brands associated with a product listing by utilizing information from both textual descriptions and accompanying images.

[0055] A brand-oriented machine learning model was trained using the Multidimensional Authenticity Analysis dataset (i.e., the item listing system data, VeRO data, and PSA data "Brand Protection and Validation Data"). The Multidimensional Authenticity Analysis dataset, used as training data, is a comprehensive collection of diverse and interconnected information sources used to train the machine learning. This dataset integrates various dimensions related to brand authenticity, legitimacy, and integrity from multiple data sources, such as item listing data and brand protection and validation data. It includes rich data sources covering product descriptions, images, seller information, customer feedback, trademark registrations, copyright applications, and legal documents. Each dimension within this dataset is represented by a set of features extracted from relevant data sources, including textual, visual, numerical, and contextual features.

[0056] VeRO (Verified Rights Owner) data refers to information provided by rights owners to online marketplaces to help protect their intellectual property. Rights owners (such as trademark or copyright owners) submit VeRO reports to inform marketplaces of lists of infringements against their intellectual property. When a rights owner identifies a list of infringements, they can submit a VeRO report to the online marketplace. This report includes details of the infringement, such as the specific listing URL, the type of infringement (e.g., trademark infringement, copyright infringement), and evidence supporting the claim. Once the marketplace receives a VeRO report, they review the information provided and can take actions such as removing the infringing listing, contacting the seller to resolve the issue, or even suspending the seller's account in the event of a repeat violation. VeRO data helps online marketplaces maintain the integrity of their platforms by addressing intellectual property infringement and ensuring a fair and secure environment for both buyers and legitimate sellers.

[0057] Post-sale verification data refers to the information and verification records collected after a transaction is completed to confirm the authenticity of the purchased item. This is especially important on e-commerce platforms selling high-value, branded, or luxury items. This data includes transaction details, expert physical inspection reports, certification certificates from recognized institutions, buyer feedback and requests, tracking and shipping records, return and refund data, and historical sales data for similar items. It ensures that the item delivered to the buyer is genuine and matches the seller's description, thus maintaining trust and credibility in the market. By verifying sold items, the platform protects buyers from counterfeit goods, increases buyer satisfaction, and maintains market integrity.

[0058] VeRO (Verified Rights Owner) data, PSA (Post-Sale Certification) data, and item listing system data are invaluable for training machine learning models to predict brands lacking brand information in item listings. Machine learning models can predict brands and further provide metadata (e.g., brand characteristics) from VeRO, PSA, and item listing system data associated with the predicted brand and any relevant brand security and item listing information. VeRO data includes verified brand information, trademark and rights ownership details, which serve as fundamental facts for brand identification. PSA data includes transaction details, physical inspection reports, certificates of authenticity, buyer feedback, tracking and shipping records, return and refund data, and historical sales data, providing a comprehensive understanding of item authenticity and brand characteristics. Item listing system data also includes transaction details, seller data, buyer feedback, tracking and shipping records, return and refund data, and historical sales data.

[0059] Features such as brand name, logo, associated trademarks, product descriptions, categories, and legal documents can be extracted from VeRO data. PSA data provides features from transaction records, inspection reports, customer feedback, and historical sales data, enabling a comprehensive understanding of brand characteristics. Item listing system data can include item titles, item descriptions, images, listing sites, prices, categories, and subcategories. By combining textual features (e.g., item descriptions, customer reviews) with visual features (e.g., item images, logos) and incorporating transactional and historical sales data, a robust feature set is created to train the model.

[0060] Convolutional Neural Networks (CNNs) are particularly effective at extracting high-level features from images, such as logo patterns, product design elements, and color schemes. Transaction data from PSA records includes details about seller identification, sales volume, pricing trends, and buyer feedback. This data is transformed into digital features representing transaction patterns, historical sales trends, and customer satisfaction levels. VeRO data provides verified brand names, trademarks, and legal ownership information. Features extracted from this data include brand ownership history, trademark registration details, and previous infringement reports, which are crucial for understanding brand authenticity and legal compliance. Contextual information, such as geographic data, time trends, and user behavior patterns, can be integrated to provide additional insights into brand authenticity and market presence. For example, geographic data may reveal the regional popularity of certain brands, while time data may indicate seasonal trends in brand sales.

[0061] VeRO data, PSA data, and item listing system data are used to train machine learning models to predict brands lacking brand information in item listings. VeRO data acts as a verified source of brand information, ensuring the model is trained on accurate and legally identified brand details. This data serves as the foundational facts for model training, enhancing the reliability and credibility of brand predictions. PSA data provides detailed insights into the post-transaction verification process, including physical inspections and customer feedback. This information ensures the identification of counterfeits and the verification of brand listing authenticity, making it a crucial component of the feature set. Item listing system data comprises structured information about products available for sale, maintained by e-commerce platforms or any similar online marketplaces. This data includes various attributes and details about each item listed on the platform, facilitating customer search, discovery, and purchase.

[0062] The integration of VeRO, PSA, and Project Listing System data introduces a wide variety of features, encompassing textual, visual, and transactional dimensions. This diversity allows the model to capture different aspects of brand characteristics, from legitimate trademarks to customer perception and sales patterns. Integrating visual features from product images (enabled by PSA data) with textual and legal features from VeRO data allows the model to more effectively detect counterfeit items. For example, discrepancies between visual logos and textual brand descriptions can indicate potential fraud. PSA data adds a contextual layer, such as transaction history and buyer behavior, which helps the model understand the brand's market context. This contextual information is crucial for making accurate brand predictions, especially when the brand name is missing or unclear. In this way, the model can predict brands and further provide metadata (e.g., brand characteristics) from VeRO, PSA, and Project Listing System data associated with the predicted brand, as well as any relevant brand security and Project Listing System information.

[0063] Machine learning algorithms, such as natural language processing techniques for text analysis, convolutional neural networks for image recognition, and supervised or semi-supervised learning for integrating these features, can be used to predict missing brand information. This dataset combines basic fact tags and annotations derived from brand protection and verification data, as well as expert judgment, to provide reference points for authenticity assessment and model training. It evolves over time to adapt to changes in market dynamics, laws and regulations, and brand-related activities, thus combining real-time updates and historical data to capture temporal trends and longitudinal patterns of brand authenticity. Scalable and flexible, the dataset can include new data sources, features, and dimensions as needed, supporting a wide variety of machine learning algorithms and analytical techniques for authenticity analysis and brand security. Overall, the multidimensional authenticity analysis dataset serves as a comprehensive and diverse resource for training machine learning models, facilitating effective brand prediction and detection of counterfeit, infringing, or misleading brand listings.

[0064] The Multidimensional Authenticity Analysis dataset can significantly enhance the training of machine learning models by leveraging its rich and diverse data sources and features to predict brands lacking brand information in item listings. The Multidimensional Authenticity Analysis dataset integrates various data sources, including item listing data (e.g., product descriptions, images, seller information) and brand protection and verification data (e.g., trademark registrations, legal documents). This diverse dataset provides a comprehensive understanding of brand characteristics and associations.

[0065] Relevant features are extracted from integrated datasets to represent different aspects of the brand. Examples include: textual features, such as keywords, phrases, and semantic patterns extracted from product descriptions, reviews, and legal documents; visual features, such as analyzing product images and logos to identify visual patterns associated with a specific brand; and contextual features, such as leveraging seller information, geographic data, and time trends to understand the context in which the brand appears.

[0066] Feature fusion techniques combine multiple features into a unified representation. This can involve cascading, weighted sums, or advanced methods such as attention mechanisms to ensure the model effectively utilizes complementary information. Feature fusion in machine learning involves combining features from multiple data sources to create a more comprehensive and informative representation for model training. In the context of predicting brands lacking brand information in an item list, feature fusion enables the integration of multiple data types, such as textual, visual, and transactional features, to improve the model's accuracy and robustness. Natural Language Processing (NLP) techniques can be used to process textual data from item lists, including product titles, descriptions, and customer reviews. These methods (such as TF-IDF (Term Frequency-Inverse Document Frequency), word embeddings (e.g., Word2Vec, GloVe), or advanced language models (e.g., BERT)) transform textual information into numerical vectors that capture semantic meaning. Image recognition techniques can be used to process visual data, such as product images and brand logos.

[0067] In implementing feature fusion, text data is cleaned and labeled, images are preprocessed before being fed into the CNN, and transaction data is normalized. Legal documents are parsed to extract relevant brand information. NLP techniques are used to extract text embeddings, CNNs are used to extract visual embeddings, and statistical methods are used to extract transaction features. Legal features are encoded based on trademark and ownership details. Features from different sources are then combined into a single feature vector. Techniques such as concatenation, weighted summation, or advanced methods (such as attention mechanisms) can be used to integrate these features. The integrated feature vector is used to train machine learning models, such as neural network models or gradient boosting models, generative AI models, and LLMs.

[0068] In context, machine learning (unsupervised learning, transfer learning, generative AI, and large language models) can be used to train machine learning models. For example, these techniques improve the ability to fuse multiple features from various data modalities, thereby enhancing the richness, robustness, and versatility of machine learning models across a wide range of tasks and domains. By leveraging each other's strengths—such as discovering latent patterns, transferring learned knowledge, generating synthetic data, and understanding complex linguistic contexts—these techniques synergistically promote more effective feature fusion and ultimately improve overall model performance.

[0069] Unsupervised machine learning models aim to find patterns and structures in data without explicit supervision or labeling of the results. They are used for tasks such as clustering (grouping similar data points together) and dimensionality reduction (reducing the number of variables considered). In the context of feature fusion, unsupervised learning can be employed to combine features from different sources or modalities into cohesive representations that capture latent patterns or relationships in the data.

[0070] Transfer learning involves using knowledge gained in solving one problem to help solve different but related problems. In machine learning, this typically means transferring knowledge from a pre-trained model (usually on a large dataset) to a new model for a different task or domain. Transfer learning can improve a model's generalization ability and its ability to perform well on new tasks by allowing the model to incorporate features learned from one domain into another, thus facilitating feature fusion.

[0071] Generative AI refers to models and techniques that create new data instances similar to the training data. This includes models such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs). In terms of feature fusion, generative AI can synthesize new features or representations that capture complex relationships between original features or between different data modalities, thereby enriching the feature space that can be used to train models.

[0072] Large-scale language models are deep learning models trained on massive amounts of text data to understand and generate human language. Examples include models such as the GPT (Generative Pre-trained Transformer) family. LLM can be used for feature fusion by extracting semantic features from text or integrating textual information with other types of data, such as images or structured data. This integration can improve the richness and contextuality of the features used by machine learning models.

[0073] In this way, technique and feature fusion can combine information from multiple sources or modalities to create a unified representation that improves model performance. Unsupervised learning techniques can help merge features by discovering patterns and relationships between different data sources. Transfer learning facilitates the integration of features learned from one domain into another, thereby improving the robustness and effectiveness of fused features. Generative AI and LLM contribute by generating new features or embeddings that capture complex dependencies and semantics, thus enriching the feature space used to train more powerful machine learning models.

[0074] A machine learning model (such as a neural network, random forest, support vector machine, generative AI model, or LLM) is trained using fused features. Even when no explicit brand information is available in the item list, the model learns to identify patterns and associations indicating specific brands. Basic fact labels from brand protection and verification data, as well as expert-annotated datasets, provide a robust foundation for supervised learning. The model can compare its predictions against these labels during training to fine-tune and improve its accuracy. The model's performance is evaluated using metrics such as accuracy, precision, recall, and F1 score. Based on these evaluations, the model is iteratively refined and retrained to improve its predictive power. Hyperparameter tuning and iterative feature selection are performed to optimize model performance. By leveraging feature fusion, a robust cube is created by integrating VeRO, PSA, and item list system data, significantly enhancing the machine learning model's ability to accurately predict brands in item lists. This comprehensive approach ensures the model can make informed predictions based on a variety of complementary information sources, thereby improving brand security and authenticity verification on e-commerce platforms. Once trained, the model can be used to predict brands in new item lists that lack explicit brand information. This helps maintain the integrity and authenticity of the brand listings on the platform.

[0075] When deployed within a project listing system, the model can automatically predict and populate missing brand information in the project list, providing suggestions to sellers or automatically updating to ensure brand accuracy. Additionally, the model can flag suspicious brands that do not match the information provided by the seller or known brand patterns, thereby enhancing fraud detection and prevention. By leveraging a multidimensional authenticity analysis dataset, the platform can significantly improve its ability to accurately predict and verify brand information, thus enhancing the overall quality and credibility of the marketplace.

[0076] Consider a scenario in an item listing system where a new item list is received without explicit brand information. The item listing system addresses this situation using a machine learning model trained on a multidimensional realism analysis dataset.

[0077] Upon receiving a list of items, the item listing system processes it using a trained machine learning model. This model leverages textual features from product descriptions, visual features from product images, and contextual features from seller information to predict missing brands. For example, if the product description mentions a specific attribute or characteristic typically associated with a particular brand, or if the product image contains a recognizable brand logo or trademark, the model can use this information to infer the brand most likely associated with the item.

[0078] Once a missing brand is identified, the platform can take proactive steps to strengthen brand security. If the predicted brand matches registered trademark or intellectual property data stored in the platform's database (e.g., VeRO data), the list will be approved and tagged with the correct brand information. However, if the predicted brand does not match any known trademark, or if there are discrepancies indicating potential counterfeiting or infringing activity, the list will be flagged for further review.

[0079] Beyond brand recognition and security enforcement, machine learning model predictions also contribute to improving the overall quality of product listings. By automatically filling in missing brand information in incomplete listings, the model enhances the completeness and accuracy of product listings on the platform. This improves the shopper experience by providing shoppers with more informative and credible product information. Furthermore, by ensuring consistency and adherence to brand information standards, the product listing system fosters trust between sellers and buyers, thereby increasing user satisfaction and engagement. This ensures that only authentic brand listings are allowed on the product listing system, thus maintaining brand security and protecting the rights of brand owners.

[0080] Advantageously, embodiments of this technical solution utilize a brand-oriented AI system associated with the item list system to support the provision of brand-oriented AI management. The brand-oriented AI system provides resources for implementing a brand-oriented machine learning model training engine, a brand-oriented security administrator engine, and a brand-oriented item management engine. The operation of the brand-oriented AI system provides a solution to problems in item list systems, such as the limited integration of AI in brand protection. The orderly combination of brand-oriented AI system components, infrastructure, and steps is an improvement over conventional item list systems, which lack support for brand protection functionality from a brand-oriented AI system.

[0081] For brand-oriented machine learning model training engines, conventional item listing systems have limited understanding of items (e.g., items or products listed in the item listing system) (especially the brand information associated with those items). Especially in the context of a product, item, or service, a brand can refer to a logo or name, which is associated with the overall perception and reputation of a product, service, or company in the minds of consumers. Therefore, brand-oriented machine learning models are trained on multidimensional authenticity analysis datasets (with specific training data used to train brand-oriented machine learning models). For example, machine learning model training can be based on metadata and unique identifiers from VeRO and PSA data, item listing system data from sellers and buyers, text, and data. Integrating these data categories allows for a holistic approach to counterfeit detection and brand authenticity assessment. This comprehensive approach significantly improves the accuracy and reliability of predictions related to counterfeit detection and brand authenticity.

[0082] For brand-oriented security administrator engines, conventional item listing systems have limited brand-oriented security measures, security planning and monitoring, and security visualization tools. Therefore, a brand-oriented machine learning model trained on a multidimensional authenticity analysis dataset is deployed within the item listing system. This brand-oriented machine learning model is integrated into the brand-oriented security administrator engine, which supports monitoring, analyzing, and enforcing brand security measures. The brand-oriented machine learning model focuses on protecting the integrity and authenticity of brands represented on the item listing system while mitigating the risks associated with counterfeiting and unauthorized brand use. The brand-oriented security administrator engine generates comprehensive reports and analyses of brand-related security metrics, including brand compliance rates, infringement incidents, enforcement actions taken, and overall item listing system integrity. The brand-oriented security administrator engine also supports visualizing brand-oriented security analytics results data by presenting complex brand-oriented security analytics results data in a clear, actionable format to enhance usability and decision-making. In this way, the brand-oriented security administrator tool leverages a novel brand-oriented machine learning model to provide novel functionalities associated with brand monitoring and detection, dashboards, real-time alerts and notifications, and trend analysis graphs.

[0083] For brand-oriented listing management engines, conventional listing systems have limited listing functionality supported by machine learning capabilities, specifically brand-oriented machine learning models. Therefore, a brand-oriented machine learning model trained on a multidimensional realism analysis dataset is deployed within the listing system. This model is integrated into the brand-oriented listing management engine, which supports listing management functions, including seller listing processes and listing quality signal analysis. The brand-oriented listing management engine ensures listing accuracy conforms to market standards and optimizes search visibility and customer experience. It also supports identifying deliberate circumvention or violation of the listing system's product listing guidelines and taking corrective actions to improve listing quality. In this way, the brand-oriented listing management tool leverages novel brand-oriented machine learning models to provide novel functionalities associated with listing management, including seller listing processes and visibility optimization based on listing quality signals.

[0084] Example systems and resources

[0085] You can use examples and refer to Figure 1A-Figure 1B To describe the various aspects of this technical solution. Figure 1A The illustration shows a project listing system 100, which includes: a project listing system with integrated artificial intelligence 100A; a brand-oriented artificial intelligence system 110, including a project listing system service 100B with an integrated API 100C; a brand-oriented application 100D; a brand-oriented artificial intelligence system resource 112; a multidimensional authenticity analysis dataset 114; a brand-oriented machine learning model training engine 120A; a brand-oriented security administrator engine 120B; a brand-oriented list management engine 120C; and a brand-oriented machine learning model 140; a project listing system client 130A; a project listing system client 130B; and a project listing system client 130C.

[0086] The project listing system 100 provides a system (e.g., a project listing system with integrated artificial intelligence 100A) that includes systems (e.g., a brand-oriented artificial intelligence system 110) and engines (e.g., a brand-oriented machine learning model training engine 120A; a brand-oriented security administrator engine 120B; a brand-oriented listing management engine 120C) for using resources (e.g., brand-oriented artificial intelligence system resources) to provide brand-oriented artificial intelligence system functionality. For example, brand-oriented artificial intelligence system resources 112 include operational, interface, and data resources. Operations include brand monitoring, sentiment analysis, product categorization, and fraud detection, ensuring continuous vigilance against brand-related violations, customer sentiment, and fraudulent activities. Interfaces include user-friendly dashboards, APIs for seamless integration, and reporting tools for administrators and brand owners to effectively monitor and manage their brand image. Data sources include brand protection and verification databases, product listings, customer feedback, and market data, providing comprehensive insights into brand protection strategies, law enforcement actions, and market trend analysis. These resources help maintain brand integrity, thereby fostering customer trust and maintaining regulatory compliance within the project listing ecosystem.

[0087] The brand-oriented AI system 110 may also include a project list system service 110B corresponding to different services of the project list system. The project list system service may include brand prediction services, brand security services, and list management services, which, for example, employ a brand-oriented machine learning model 140 to provide project list system functionality. The project list system service 100B may include an integration application programming interface (API) 110C for integrating the project list system service with the brand-oriented machine learning model 140. The brand-oriented application 110D may also operate to employ the brand-oriented machine learning model 140 to provide functionality associated with the generative AI application 110D. The embodiments described herein contemplate other variations and combinations of the project list system service.

[0088] The listing system 100 may also include listing system clients (e.g., listing system client 130A (administrator client), listing system client 130B (seller client), and listing system client 130C (buyer client)). Administrators responsible for overseeing the functionality of the listing system 100 use listing system client 130A to configure and maintain the listing system 100, establish categories and attributes, and monitor user activity to ensure compliance with platform policies. Sellers (as part of the platform's product offerings) utilize listing system client 130B to create and manage listings, update product details, pricing, and inventory levels, while processing orders and handling customer inquiries. Interacting through a dedicated dashboard, sellers proactively manage their storefronts, adjust listings, and communicate seamlessly with buyers. Buyers (end-users of the platform) rely on listing system 130C via listing system client 130C to explore, evaluate, and purchase products. They browse product listings, filter results, and view details through the system's intuitive interface before making a purchase decision. Through interaction with the platform, buyers seamlessly complete transactions, track orders, and receive updates throughout the process. The project list system client facilitates client functionality associated with the brand-oriented AI system 110.

[0089] Reference Figure 1B , Figure 1B It shows having Figure 1A The system includes a project list system 100, and also includes brand protection and verification data 114A, project list data 114B, security data 150, and project list 160.

[0090] Brand-oriented machine learning model training engine

[0091] A brand-oriented machine learning model training engine 120A accesses a multidimensional authenticity analysis dataset 114 associated with multiple brands and uses the multidimensional authenticity analysis dataset 114 to train a brand-oriented machine learning model 140. The multidimensional authenticity analysis dataset 114 includes intellectual property enforcement data, certified souvenir data, and item listing system data (e.g., brand protection and verification data 114, and item listing system data 114B), which provide basic fact labels and annotations associated with model training.

[0092] The brand-oriented machine learning model 140 is trained based on multidimensional brand authenticity features. This training utilizes machine learning techniques that integrate multidimensional brand authenticity features corresponding to intellectual property enforcement data, certified souvenir data, and item list system data. The brand-oriented machine learning model 140 is trained by receiving item list information (e.g., item list 160) without brands as input for generating predicted brands; this item list information serves as an indicator with features relevant to generating the predicted brands.

[0093] A brand-oriented machine learning model training engine 120A deploys a brand-oriented machine learning model 140 in an item list system to support one or more applications (e.g., a brand-oriented application 100D). These applications include brand-oriented security administrator tools and brand-oriented list management tools. The brand-oriented machine learning model 140 is configured to generate predicted brands for item lists that do not include brands, wherein the predicted brands are generated based on item list information associated with the item lists. The brand-oriented machine learning model 140 is a multimodal model that includes a text-based model and an image recognition model. A first predicted brand score is associated with the text-based model, and a second predicted brand score is associated with the image recognition model, such that the generated predicted brands are based on a combination of the first and second predicted brand scores.

[0094] Brand-oriented security administrator engine

[0095] The Brand-Oriented Security Administrator Engine 120B accesses Brand-Oriented Security Data 150. Brand-oriented can be a list of items that does not include brand information, or a combination of lists with and without brand information. Access to Brand-Oriented Security Data is based on a security administrator's request for Brand-Oriented Security Analytics Results Data, which is associated with a brand identifier in the request. The brand identifier indicates that Brand-Oriented Security Analytics Results Data should be generated for the brand identifier.

[0096] The brand-oriented security administrator engine 120B uses a brand-oriented machine learning model 140 trained on a multidimensional authenticity analysis dataset 114 and brand multidimensional authenticity features to analyze brand-oriented security data 1250. The brand-oriented machine learning model 140 is integrated into a brand-oriented security administrator tool, which uses the brand-oriented security administrator engine 120B of the item list system 100 to support the generation of brand-oriented security analysis results data.

[0097] The Brand-Oriented Security Administrator Engine 120B generates brand-oriented security analytics results data. The generation of brand-oriented security analytics results data is based on: identifying predicted brands from a list of items, where the list does not include brand input; extracting item list information associated with the item list; and updating existing brand-oriented security analytics results data for predicted brands based on the extracted item list information.

[0098] The Brand-Oriented Security Administrator Engine 120B can also generate brand security strategies based on brand-oriented security analytics data. Generating brand strategies based on brand-oriented security analytics involves analyzing data on security incidents, reputation, and customer sentiment. Specifically, generating brand strategies can include identifying threats, assessing brand perception, and defining strategic objectives. This process includes developing communication plans, implementing security measures, and monitoring performance.

[0099] The Brand-Oriented Security Administrator Engine 120B delivers brand-oriented security analytics results data, enabling the display of one or more visualizations based on these results. The Brand-Oriented Security Administrator Engine 120B can create security incident trend charts or geospatial event heatmaps. For example, the visualization displays a trend chart showing the frequency and distribution of security incidents over time. Each incident type (e.g., data breach, phishing attack, malware infection) is represented by a different color or symbol on the chart. The x-axis represents time (e.g., days, weeks, months), while the y-axis shows the number of incidents. This chart helps visualize the overall trends of security incidents affecting a brand and identify periods of increased activity or vulnerabilities.

[0100] Another visualization presents a geographic heatmap, showing the geographical distribution of security incidents. Each incident is plotted on the map, with color intensity indicating the concentration or severity of the incident in different regions. Stakeholders can zoom in on specific areas to identify hotspots of security activity or vulnerabilities. This visualization helps to understand the geographic distribution of security threats and prioritize mitigation measures for high-risk areas.

[0101] Brand-oriented list management engine

[0102] The brand-oriented list management engine 120C accesses the item list associated with the item list system and analyzes the item list using a brand-oriented machine learning model 140 trained based on a multidimensional authenticity analysis dataset 114 and brand multidimensional authenticity features. The brand-oriented machine learning model 140 is integrated into a brand-oriented list management tool that uses the brand-oriented list management engine 120C of the item list system 100 to support brand-oriented list management.

[0103] Based on the analyzed list of items, the Brand-Oriented List Management Engine 120C generates brand-oriented security notifications associated with the list. The generation of brand-oriented security notifications is based on identifying predicted brands within the list of items, where the list does not include brand input. A brand-oriented security notification is a communication or alert specifically designed to inform stakeholders, such as customers, partners, and employees, of security incidents or threats that directly impact a brand's reputation, credibility, or integrity.

[0104] The Brand-Oriented Listing Management Engine 120C sends Brand-Oriented Security Notifications. These notifications are sent to the sellers associated with the listings and the security administrators, administrators, or another service of the Brand-Oriented Listing Management Engine 120C. Based on these notifications, the Brand-Oriented Listing Management Engine 120C performs listing quality analysis, in part based on the Brand-Oriented Security data associated with them. Listing quality analysis involves a comprehensive evaluation of the listings, covering various attributes such as titles, descriptions, images, prices, and reviews.

[0105] Listing quality analysis can focus on fraud, involving a meticulous examination of item listings to identify potential fraudulent activity or listings. This comprehensive assessment includes carefully reviewing various attributes such as product titles, descriptions, images, prices, and seller information. Data collection involves gathering information from the listings and analyzing it for consistency, irregularities, or red flags that may indicate fraudulent activity. Attributes such as excessively low prices, misleading descriptions, or suspicious seller profiles are flagged for further investigation. Content analysis delves into the accuracy and authenticity of product information to ensure listing transparency and compliance with platform policies. Image quality assessment verifies the legitimacy of product images and their correspondence with listed items. Pricing analysis includes comparing prices to market averages and identifying anomalies that suggest fraudulent pricing practices.

[0106] Additionally, seller reputation and past transaction history are examined to assess credibility and detect patterns of fraudulent activity. Based on the analysis, recommendations for fraud prevention and mitigation strategies are then generated, focusing on strengthening the listing verification process, implementing more stringent seller vetting processes, and enhancing fraud detection algorithms. Continuous monitoring and iteration are performed to adapt to evolving fraud tactics and ensure the integrity and security of the listing system. Based on listing quality analysis, the listing can be flagged for remedial action.

[0107] In one embodiment, listing management may include restricting the visibility of listings until brand information is confirmed or corrected. Listings with uncertain or unverified brand details may be temporarily hidden from the general view, prompting sellers to update and verify the information promptly. Initially, if a listing lacks verified brand details, these listings are flagged and hidden from the general view, prompting sellers to update and confirm the information promptly through the user interface. Machine learning models predict brand information based on listing attributes such as descriptions and images, verifying or suggesting corrections. Sellers are notified to review and modify brand information, triggering a reassessment and update of the visibility status once verified.

[0108] Turning Figure 2A , Figure 2A Example item list interface 200A is shown, associated with missing brand information. The item list interface 200A for wallet item 202A features an image showing wallet design details and a descriptive title 206A specifying style and any unique features. Price and availability information is presented in item list interface section 208A, allowing users to assess the item's value and inventory status. Item list interface section 210A includes item details and descriptions from the seller. However, brand information is missing and listed as "No Brand." Brand information may have been unintentionally or intentionally omitted. Therefore, a brand-oriented AI system could be used to analyze this type of item list to predict brands, implement brand security strategies, or manage item lists.

[0109] Turning Figure 2B , Figure 2BAn example brand-oriented AI system associated with providing brand-oriented AI management is illustrated according to embodiments described herein. At a high level, input 210 is processed and analyzed via a machine learning (ML) model 230B (e.g., a brand-oriented machine learning model). Input 210 serves as the basis for the ML model 230B to make predictions to generate output 240 (e.g., predicting brands). The brand-oriented AI system integrates different types of applications (or services) that take the information from output 240. Input 210 may include list titles, list sites 214B, categories 216B, item descriptions 218B, list images 220, feedback 222B, return messaging 224B, machine-to-machine (M2M) messages, and item details 228B. Each input serves as an indicator for the ML model 230 to predict the brand of an unbranded list of items associated with input 210. For example, list site 214B can employ customized methods to evaluate fraud patterns, market dynamics, payment methods, language and cultural factors, and geolocation data. Fraudulent activity often varies based on geolocation. Due to cultural factors, economic conditions, or technological infrastructure, certain types of fraud or scams may be more prevalent in specific regions. Analyzing regional fraud patterns can help develop fraud detection strategies to address specific risks associated with brands.

[0110] ML model 230B may include word embedding model 232B and image recognition model 234B. It is conceivable that ML model 230B can operate independently or in combination, or as a multimodal model to provide brand prediction functionality. For example, the two models could operate independently, each producing its own predicted brand score for a given list of items. However, to improve accuracy and reliability, ensemble methods (such as voting ensemble, where each model's predicted brand score is treated as a "vote" for the predicted brand) can be used to combine the predicted brand scores from the two models. The final prediction is determined by aggregating these votes (such as taking the average or weighted average of the scores). This method leverages the collective insights of multiple models to make more informed decisions. Stacked ensemble uses the predicted brand scores from both models as features to train a meta-model, which learns how best to combine these scores to make the final prediction. This meta-model effectively learns the strengths and weaknesses of each base model and optimally integrates its predictions to improve overall performance.

[0111] Beyond generating predicted brand scores to anticipate missing brands within an item list, various types of information can be generated and integrated into the output 240B to enhance the accuracy and depth of the analysis. In addition to the predicted missing brands themselves, key contextual details such as product descriptions, categories, and price ranges provide valuable insights into potential brand associations. Analysis of similar products, user reviews or ratings, and seller information further enrich the understanding by elucidating common brand mentions, sentiment, and seller-brand relationships. Image recognition technology can provide visual cues from product images, helping to identify logos or trademarks associated with specific brands. Furthermore, access to external databases or repositories of brand information allows for comprehensive comparison and validation of predicted brands against established brand directories. By integrating these diverse information sources, machine learning models can generate more accurate predictions, ultimately contributing to improved decision-making and data-driven insights in the field of brand identification and analysis within item lists.

[0112] Output 240B can be transmitted to the application-related proactive counterfeit detection 252B, listing quality signals 254B, reporting insights and analysis 256B, and seller listing process 258B. Proactive counterfeit detection 252B can include practices that proactively identify and prevent the sale of counterfeit or fraudulent items before they are listed or sold on the platform. It involves processes such as detecting suspicious listings, verifying product authenticity, and taking preemptive action to remove or block fraudulent sellers or listings. Listing quality signals 254B are indicators or metrics used to evaluate the quality, accuracy, and reliability of listings on the platform. Listing quality signals 254B can include factors such as detailed product descriptions, high-quality images, accurate categorization, competitive pricing, positive user reviews, and seller reputation. Monitoring and analyzing these signals helps the platform ensure a positive user experience, maintain trust, and reduce risks such as counterfeit or misleading listings.

[0113] Reporting insights and analytics (256B) involve collecting, analyzing, and deriving actionable insights from reports submitted by users, moderators, or automated systems regarding various aspects of platform activity, such as suspicious behavior, policy violations, or user feedback. Reporting insights and analytics enable the platform to identify trends, patterns, and areas of concern, make informed decisions, and implement targeted interventions to address issues and improve platform integrity and user satisfaction.

[0114] The Seller Listing Process 258B refers to the step-by-step process sellers follow to create and publish listings on the platform. The Seller Listing Process 258B typically includes tasks such as entering product details, uploading images, setting prices and inventory information, selecting categories or tags, and submitting the listing for review or publication. Optimizing the Seller Listing Process involves streamlining the process, providing guidance or tools to improve listing quality, and implementing checks or verifications (e.g., missing brand information) to prevent errors or policy violations, ultimately improving the efficiency and effectiveness of sellers' activities on the listing platform.

[0115] Turning Figure 2C , Figure 2C An example brand-oriented machine learning model flow associated with generating predicted brands according to embodiments described herein is illustrated. For example, the brand-oriented machine learning model may be a word embedding model serving as input 202C, and performing hashing 204C, averaging 206C, linear transformation 208C, and softmax 210C operations on preprocessed comments 212C to generate word vectors 214C, hidden comment representations 215C, transformed comment vectors 218C, and predicted brand output 220C.

[0116] Input preprocessing: Comments are preprocessed (e.g., L1 preprocessing of comments) to remove punctuation, convert text to lowercase, and mark words.

[0117] Hash 204C: Hashes are used to map each word to a fixed-size vector representation. Instead of using a traditional lookup table for word embedding, a hash function is applied to convert each word into a fixed-size hash value. This ensures efficient memory usage and scalability, especially for large vocabularies.

[0118] Calculate the average 206C: Average the hashed word vectors (e.g., L2 N word vectors) to obtain a single vector representation of the entire comment. This average vector captures the overall semantics of the comment by aggregating information from the individual word embeddings.

[0119] Hidden comment representation 210C: Hidden comment vectors (e.g., L3 hidden comment representations) serve as hidden representations of comments. This vector encapsulates the learned features and semantic information extracted from the comments, which is crucial for sentiment analysis.

[0120] Linear Transformation 208C: A linear transformation is applied to the average vector to project it onto a higher-dimensional space (e.g., an L4-transformed comment vector). This transformation helps capture complex relationships and patterns in comment data that may not be captured by the original word embeddings.

[0121] Predicted Output: Finally, the hidden comment representation is fed into a classifier (such as a softmax layer) to predict the brand of the comment (e.g., GUCCI, COACH, or LOUIS VUITTON). The softmax operation calculates the probability of each possible sentiment label based on the learned features. Each brand label corresponds to a category, and the classifier outputs the predicted brand label with the highest probability as the predicted brand for the item list. The classifier's output provides the predicted brand label based on the learned features and representations extracted from the comment data.

[0122] Turning Figure 3 , Figure 4 , Figure 5 and Figure 6 These diagrams illustrate corresponding interfaces and illustrations related to brand security. For example, these interfaces can be associated with security administrator tools that allow reviewing details of individual item lists and their corresponding details. (See reference...) Figure 3 , Figure 3 Chart 302 is shown, providing a comprehensive overview of missed events occurring throughout the year, focusing on, for example, December, April, and August. The x-axis 304 of this chart represents the progression of time divided by the months of the year (i.e., December, April, and August). Simultaneously, the y-axis 306 quantifies the total number of missed events recorded during each corresponding month. For each month, data points are plotted on the chart to represent the total number of observed missed events. These data points are interconnected to form lines (e.g., 310 and 320), thus creating distinct visual representations for December, April, and August.

[0123] In addition to showing monthly omission event data, the chart includes dashed lines to divide it into two components: the annual target omissions (e.g., 310) and the historical baseline (e.g., 310). The dashed line indicating the annual target omissions 310 serves as a benchmark or predetermined target for measuring performance. These lines extend horizontally across the chart, intersecting the data points for each month. They provide a clear visual reference, allowing stakeholders to assess whether the total omissions for a given month are consistent with the target established for that year. Furthermore, the chart is characterized by a historical baseline 320 represented by another dashed line. This baseline reflects the average or historical trend of omissions observed throughout the year. It serves as a comparison point, allowing viewers to assess current performance relative to past patterns.

[0124] In this context, "bps" stands for baseline, a unit of measurement used to quantify small percentage changes. In the context of this chart, bps can indicate deviation from a historical baseline or target omissions expressed as percentage points. For example, if an actual omission exceeds the baseline or fails to reach a target, the deviation can be quantified in bps to provide a precise understanding of the magnitude of the difference. Overall, this chart provides a comprehensive visualization of omissions, facilitating informed decision-making and strategic planning by offering insights into performance relative to targets and historical trends.

[0125] Reference Figure 4 , Figure 5 and Figure 6 , Figure 4 An example dashboard interface 402 is shown, in which charts (e.g., charts 410 and 420) are associated with a total list of suspicious items identified using proactive methods (e.g., y-axis 412) and a total list of suspicious items identified using passive methods (e.g., y-axis 422). The charts can be plotted weekly (e.g., x-axis 412 and x-axis 422). For example, chart 410 provides a visual representation of the total number of suspicious items identified weekly using proactive methods. The x-axis of this chart represents the progress of time, divided into weeks, while the y-axis quantifies the total count of suspicious items. Proactive methods for detecting actual or potential (“suspicious” fraud lists involve employing preemptive strategies and advanced algorithms to identify potential fraud instances before they escalate. These methods typically involve data analysis, machine learning algorithms, and pattern recognition techniques to detect anomalous or suspicious patterns in list behavior. Each data point on chart 410 represents the cumulative count of suspicious items identified using proactive methods during a specific week. These data points are interconnected to form a line graph, which illustrates the trend of proactive detection efforts over time. This chart can show the fluctuations in the number of suspicious items identified each week, thus reflecting changes in the effectiveness of proactive detection measures or changes in fraudulent activity patterns.

[0126] Turning to the second chart 420, which shows the list of suspicious items identified using a passive method, and chart 430 depicting the total number of suspicious items identified by the passive method on a weekly basis. Similar to chart 410, the x-axis represents time in weeks, while the y-axis indicates the total count of suspicious items.

[0127] Passive methods for detecting actual or potential (“suspicious” lists involve responding to reported instances of suspicious activity or anomalies identified through customer complaints, manual review, or post-transaction monitoring. These methods rely on human intervention and investigative techniques to address fraudulent activity after it has already occurred. Each data point on Chart 420 represents the cumulative count of suspicious items identified by passive methods during a specific week. Similar to Chart 410, these data points are connected to form a trend line, which shows the trajectory of passive detection efforts over time. This chart can reveal fluctuations in the number of suspicious items identified by passive methods, reflecting changes in the number of reported incidents, the efficiency of the investigative process, or changes in consumer behavior patterns. Overall, these two charts provide valuable insights into the effectiveness of both active and passive methods for detecting suspicious items, enabling stakeholders to assess the effectiveness of their fraud detection strategies and make informed decisions to mitigate risk.

[0128] Reference Figure 5 and Figure 6 These figures illustrate dashboard interface 502 and instruments and dashboard 602. Dashboard interface 502 includes a first interface portion 510 that provides selectable filters (filters 512 and 514) that can be used to filter data associated with the dashboard interface. Dashboard interface 602 includes a second interface portion 610 that provides sources (source 3 612 and source 9 614) that indicate corresponding counts or percentages of active and / or passive items associated with corresponding sources. The dashboard interface can be associated with different types of brand security features, such as brand monitoring, list analytics, geographic analytics, user feedback and reporting, interface features, and graphics and charts.

[0129] For example, dashboards can display dynamic trend graphs that further clarify the fluctuation patterns of fraudulent activity over time, enabling administrators to identify emerging trends and respond accordingly. Dashboards facilitate vigilant brand monitoring, a key component of fraud detection and prevention. They present administrators with an intuitive interface allowing them to manage lists of monitored brands, easily adding or removing monitored entities. Seamless integration of real-time alerts immediately notifies administrators of any suspicious listings associated with monitored brands, ensuring swift intervention to protect brand integrity.

[0130] Furthermore, the dashboard supports the presentation of data associated with the analytics of each listing flagged as fraudulent. Each flagged listing is meticulously categorized and accompanied by thumbnails for quick visual identification. Relevant details such as seller information, listing date, price, and location are carefully categorized, providing administrators with a comprehensive understanding of current fraudulent activity. Administrators can delve deeper into each listing, accessing a wealth of detailed information and actionable insights, enabling them to take decisive actions such as removing the listing or blocking the seller.

[0131] The dashboard provides geographic visualization, presenting administrators with an interactive map interface depicting the spatial distribution of the fraud list. Dynamic heatmaps overlay this map vividly show areas with the highest density of fraudulent activity. Administrators can seamlessly delve into specific regions or cities, facilitating targeted interventions to curb fraudulent activity at its source.

[0132] The dashboard includes robust user feedback and reporting mechanisms, enabling users to report suspicious listings or activities, thus contributing to the collective effort to combat fraud. A dedicated section supports tagging potential fraudulent activity for administrative review. Tagged lists are highlighted to ensure they immediately reach administrators' attention. Clicking on a tagged list reveals a wealth of detail, allowing administrators to gain a deeper understanding of the complexities of each situation. Every aspect of the listing, from product descriptions to seller history, is meticulously documented, enabling administrators to make informed decisions. Furthermore, the interface features seamless communication tools, allowing administrators to interact directly with sellers, fostering dialogue and driving solutions. An effective escalation mechanism is in place to ensure serious incidents are promptly reported to the appropriate authorities for further action. Administrators can leverage the power of data analytics to analyze the fraudulent item list, identifying recurring patterns or emerging trends to strengthen fraud prevention measures.

[0133] Additionally, the dashboard provides access to a wealth of graphs and charts, transforming raw data into actionable insights. Time-series graphs track the trajectory of fraudulent activity over time, enabling administrators to identify temporal patterns and predict future trends. Pie charts depict the distribution of fraudulent listings across various product categories, revealing areas requiring enhanced scrutiny. Bar charts highlight top brands involved in fraudulent activity, facilitating targeted interventions to maintain brand integrity. Finally, interactive heatmaps provide a spatial perspective, highlighting areas where rampant fraudulent activity is being addressed, guiding administrators to strategically deploy resources to effectively combat fraud. Essentially, the dashboard enables the implementation of brand security measures linked to brand-oriented machine learning models, providing administrators with the tools and insights needed to maintain the integrity of the item listing system and protect the interests of users and stakeholders.

[0134] Turning Figure 7A , Figure 7B , Figure 7C , Figure 8 , Figure 9A and Figure 9B These diagrams illustrate the corresponding interfaces associated with the list of items. For example, these interfaces could be associated with a security administrator tool that allows reviewing the details of an individual list of items and its corresponding details. Figure 7A , Figure 7B and Figure 7C The results show a true positive scenario where the brand-oriented machine learning model correctly predicts the correct brand. Figure 8 An example of a false positive was shown, and Figure 9A and Figure 9B An abnormal situation was indicated.

[0135] A true positive occurs when the model correctly identifies a list as belonging to a specific brand and the prediction is indeed accurate after verification. Essentially, it indicates that the model correctly detected the presence of a brand it was designed to identify. A false positive occurs when the model incorrectly predicts the presence of a specific brand in the list, but after verification, the list does not actually belong to that brand. Anomalies are instances or situations that deviate significantly from the normative or expected pattern within a given context (e.g., item list information). These situations typically exhibit characteristics, behaviors, or outcomes that are uncommon, unexpected, or irregular compared to typical occurrences in the considered domain. False positives and anomalies can be alternative handling mechanisms for brand safety. For example, these types of situations can be escalated, such as escalating instances marked as outliers or anomalous to a higher level of review or decision-making within the organization. This may require sending anomalous predictions to domain experts, senior data scientists, or management for further analysis, verification, or intervention. The goal is to ensure that these situations receive appropriate attention and action, including human review, model refinement, or process improvement, to enhance the accuracy and reliability of the brand prediction system. The embodiments described herein envision other variations and combinations of handling false positives and anomalies.

[0136] Finally, the confidence score (i.e., the predicted brand score) represents the level of certainty or confidence the model has in its predictions. It indicates the model's belief in the correctness of its classification output. Confidence scores typically range between 0 and 1, where 0 indicates low confidence and 1 indicates high confidence. For example, if a model assigns a confidence score of 0.8 to a prediction, it means that the model is 80% confident in its classification decision.

[0137] Reference Figure 7A , Figure 7B and Figure 7CAs discussed, the input data associated with the project can be used to generate predicted brands. For illustration, a sample project list information interface 702 is provided. We will discuss the seller tags shown in the seller tag interface section 710, including the brands provided by the seller and the images in the image section 720 of the interface. Figure 7A In the list, brand 702A is "unbranded," and processing this item list using a brand-oriented machine learning model predicts that the brand is "Brand A," with a predicted brand score of 0.99996. This prediction can be based on seller tag 704A (e.g., faux leather, monogram, tweed) and / or the corresponding image 708A. Figure 7B In the data, brand 702B is "unbranded," and processing the item list can predict this brand as "Brand B," with a predicted brand score of 1.0. This prediction can be based on seller tag 704B (e.g., leather, crossbody bag) and / or the corresponding image 706B. Figure 7C In the list, brand 702C is "unbranded," and processing the item list can predict the brand as "brand C," with a predicted brand score of 0.99494. This prediction can be based on seller tag 704C (e.g., HD03, 1600W, professional hair dryer) and / or the corresponding image 706C.

[0138] Reference Figure 8 This indicates a false positive. Figure 8 In the list, brand 802C is "many", and processing the item list can predict the brand as "brand X", where the predicted brand score is 0.800098. This prediction can be based on seller tag 804C (e.g., paper cloth, all occasions, gift tote bags) and / or the corresponding image 806C.

[0139] Reference Figure 9A and Figure 9B These solutions address the abnormal situations. Figure 9A This includes the seller's brand 902A "Brand X", but the product line in seller tag 904A is "Brand Y" and an image of a Brand X wallet. The generated predicted brand could be "Brand X", with a predicted brand score of 0.9992. However, this information conflicts with the product line. Therefore, this may be a candidate for further review (e.g., item listing quality analysis and notifications to update seller tags).

[0140] Figure 9B This includes brand 902B provided by the seller, "Brand A, Brand X, Brand Y, Brand Z," and several seller tags (e.g., mixed, leather mixed, mixed), with the predicted brand being "Brand Z," which has a predicted brand score of 0.3422. Therefore, this could be a candidate for further review.

[0141] You can use examples and refer to Figure 1A and Figure 1B This is how we describe various aspects of a technical solution. Figure 1A Based on reference Figures 19-21 A block diagram of an exemplary technical solution environment for implementing embodiments of the technical solution is described. Typically, this technical solution environment includes a technical solution system suitable for providing an example item list system 100 that can employ the methods of this disclosure. Specifically, Figure 1A A high-level architecture of an item list system 100 according to an embodiment of this disclosure is shown. In addition to other engines, managers, generators, selectors, or components not shown (collectively referred to herein as "components"), Figure 1A Project list platform system 100 corresponds to Figure 1B .

[0142] Example Method

[0143] Reference Figures 10-18 Flowcharts illustrate methods for providing brand-oriented AI management capabilities within a project listing system. These methods can be performed using the project listing system described herein. In embodiments, one or more computer storage media have computer-executable or computer-usable instructions that, when executed by one or more processors, cause the one or more processors to perform methods (e.g., computer-implemented methods) within a project listing platform system (e.g., a computerized system or computer system).

[0144] Turning Figure 10 A flowchart is provided illustrating method 1000 for providing brand-oriented AI management capabilities in a project listing system. At box 1002, a multidimensional authenticity analysis dataset associated with multiple brands is accessed; at box 1004, the multidimensional authenticity dataset is used to train a brand-oriented machine learning model; and at box 1006, the brand-oriented machine learning model is deployed in the project listing system to support one or more applications.

[0145] Turning Figure 11A flowchart is provided illustrating method 1100 for providing brand-oriented AI management capabilities in an item listing system. In box 1102, a multidimensional authenticity analysis dataset associated with multiple brands is accessed; in box 1104, a word embedding model of the brand-oriented machine learning model is trained using the multidimensional authenticity dataset; in box 1106, an image recognition model of the brand-oriented machine learning model is trained using the multidimensional authenticity dataset; and the brand-oriented machine learning model is deployed in the item listing system, which selectively employs the word embedding model and the image recognition model to predict the brands of the item listings.

[0146] Turning Figure 12 A flowchart is provided illustrating method 1200 for providing brand-oriented AI management capabilities in a project listing system. At box 1202, the process includes: identifying a multidimensional authenticity analysis dataset; selecting one or more machine learning algorithms to train a brand-oriented machine learning model; and deploying the brand-oriented machine learning model to operate in conjunction with one or more applications within the project listing system.

[0147] Turning Figure 13 A flowchart is provided illustrating method 1300 for providing brand-oriented AI management functionality in a project listing system. At box 1302, brand-oriented security data is accessed; at box 1304, the brand-oriented machine learning model trained on a multidimensional realism analysis dataset is used to analyze the brand-oriented security data; at box 1306, brand-oriented security analysis result data is generated; and at box 1308, the brand-oriented security analysis result data is transmitted to display one or more visualizations based on the brand-oriented security analysis result data.

[0148] Turning Figure 14 A flowchart is provided illustrating method 1400 for providing brand-oriented AI management functionality in a project list system. At box 1402, predicted brands for the project list are identified, excluding brand inputs; at box 1404, project list information associated with project list updates is extracted; and at box 1406, existing brand-oriented security analysis results data for predicted brands are updated based on the extracted project list information.

[0149] Turning Figure 15A flowchart is provided illustrating method 1500 for providing brand-oriented AI management functionality in a project listing system. At box 1502, brand-oriented security data is transmitted; at box 1504, brand-oriented security analysis data is received, which is generated using a brand-oriented machine learning model that analyzes the brand-oriented security data. The brand-oriented machine learning model is trained based on multidimensional realism analysis data; at box 1506, the brand-oriented security analysis results data are displayed.

[0150] Turning Figure 16 A flowchart is provided illustrating method 1600 for providing brand-oriented AI management functionality in a project list system. At box 1602, the project list associated with the project list system is accessed; at box 1604, the project list is analyzed using a brand-oriented machine learning model trained on a multidimensional realism analysis dataset; at box 1606, based on the analyzed project list, a brand-oriented security notification associated with the project list is generated; and at box 1608, the brand-oriented security notification is delivered.

[0151] Turning Figure 17 A flowchart is provided illustrating method 1700 for providing brand-oriented AI management capabilities in a project list system. At box 1702, brand-oriented safety notifications associated with the project list are accessed; at box 1704, list quality analysis is performed, in part based on brand-oriented safety data associated with the brand-oriented safety notifications; and at box 1706, based on the project list quality analysis, the project list is flagged for remedial action.

[0152] Turning Figure 18 A flowchart is provided illustrating method 1800 for providing brand-oriented AI management functionality in an item list system. At box 1802, a request to generate an item list is transmitted; at box 1804, a brand-oriented safety notification generated using a brand-oriented machine learning model that analyzes the item list is received. The brand-oriented machine learning model is trained based on a multidimensional realism analysis dataset; at box 1806, the brand-oriented safety notification is displayed.

[0153] Additional support for the detailed description of the invention

[0154] Example project list system environment

[0155] Now refer to Figure 19 , Figure 19 A list of example projects, including System 1900 computing environments, that can be implemented using the embodiments of this disclosure are shown. Specifically, Figure 19The high-level architecture of platform 1910, a list of example projects that can host a technology solution environment or a portion thereof, is shown. It should be understood that this and other arrangements described herein are illustrated as examples. For instance, as mentioned above, many of the elements described herein can be implemented as discrete or distributed components or combined with other components, and implemented in any suitable combination and location. Other arrangements and elements (e.g., machine, interface, function, command, and function groupings) may be used in addition to or in lieu of the arrangements and elements shown.

[0156] Project listing system 1900 may be a cloud computing environment that provides computing resources for functions associated with project listing platform 1910. For example, project listing system 1900 supports the delivery of computing components and services, including servers, storage, databases, networks, applications, and machine learning associated with project listing platform 1910 and client device 1920. Multiple client devices (e.g., client device 1920) include hardware or software for accessing resources on project listing system 1900. Client device 1920 may include applications (e.g., client application 1922) and interface data (e.g., client application interface data 1924) that support client functions associated with the project listing system. Multiple client devices may access the computing components of project listing system 1900 via a network (e.g., network 1930) to perform computing operations.

[0157] The 1910 project listing platform is responsible for providing a computing environment or architecture, which includes the infrastructure to support the platform's functionalities, such as e-commerce features. The platform supports storing projects in a project database and provides a search system for receiving queries and identifying search results based on those queries. The platform can also provide a computing environment with features for managing, selling, purchasing, and recommending different types of projects. The 1910 platform can be specifically designed for content platforms, such as the eBay content platform or e-commerce platform developed by eBay Inc. in San Jose, California.

[0158] The project listing platform 1910 can provide project listing operations 1930 and project listing interfaces 1940. Project listing operations 1930 may include service operations, communication operations, resource management operations, security operations, and fault-tolerant operations that support specific tasks or functions within the project listing platform 1910. Project listing interfaces 1940 may include service interfaces, communication interfaces, resource interfaces, security interfaces, and management and monitoring interfaces that support functions between project listing platform components. Project listing operations 1930 and project listing interfaces 1940 enable communication, coordination, and seamless operation of the project listing system 1900.

[0159] By way of example, the functionalities associated with the 1910 item listing platform may include: shopping operations (e.g., product search and browsing, product selection and shopping cart, checkout and payment, and order tracking); user account operations (e.g., user registration and authentication, and user profiles); seller and product management operations (e.g., seller registration, product listing, and inventory management); payment and financial operations (e.g., payment processing, refunds, and returns); order fulfillment operations (e.g., order processing and fulfillment, and inventory management); customer support and communication interfaces (e.g., customer support chat / email and notifications); security and privacy interfaces (e.g., authentication and authorization, payment security); recommendation and personalization interfaces (e.g., product recommendations and customer reviews and ratings); analytics and reporting interfaces (e.g., sales and inventory reports, and user behavior analytics); and APIs and integration interfaces (e.g., APIs for third-party integration).

[0160] The project listing platform 1910 can provide a project listing platform database (e.g., project listing platform database 1950) to effectively manage and store different types of data. The project listing platform database 1950 may include relational databases, NoSQL databases, search databases, cache databases, content management systems, analytics databases, payment gateway databases, customer relationship management databases, log and error databases, inventory and supply chain databases, and multi-channel databases, which are used in combination to effectively manage data and provide users with an e-commerce experience.

[0161] The Project Listing Platform 1910 supports applications (e.g., Application 1960), which are computer programs, software components, or services that serve a specific function or set of functions to meet the specific requirements of the Project Listing Platform or user requirements. Applications can be client-side (user-facing) and server-side (back-end). Applications can also include applications without any AI support (e.g., Application 1962), applications supported by traditional AI models (e.g., Application 1964), and applications supported by generative AI models (e.g., Application 1966). By way of example, applications can include online storefront applications, mobile shopping applications, application and management consoles, payment gateway integrations, user account and authentication applications, search and recommendation engines, inventory and stock management applications, order processing and fulfillment applications, customer support and communication tools, content management systems, analytics and reporting applications, marketing and promotion applications, multi-channel integration applications, logging and bug tracking applications, customer relationship management (CRM) applications, security applications, and APIs and web services, which are combined to effectively provide users with an e-commerce experience.

[0162] The project list platform 1910 may include machine learning engines (e.g., Machine Learning Engine 1970). Machine Learning Engine 1970 refers to a machine learning framework or platform that provides the infrastructure and tools for designing, training, evaluating, and deploying machine learning models. Machine Learning Engine 1970 can serve as the backbone for developing and deploying machine learning applications and solutions. Machine Learning Engine 1970 can also provide tools for visualizing data and model results, as well as interpreting model decisions to understand how the model makes predictions.

[0163] Machine Learning Engine 1970 provides the necessary libraries, algorithms, and utilities to perform various tasks within a machine learning workflow. This workflow can include data processing, model selection, model training, model evaluation, hyperparameter tuning, scalability, model deployment, inference, integration, customization, and data visualization. Machine Learning Engine 1970 can include pre-trained models for various tasks, simplifying the development process. In this way, Machine Learning Engine 1970 streamlines the entire machine learning process, from data preparation and model training to deployment and inference, making it accessible and efficient for different types of users engaged in a wide range of machine learning applications—e.g., clients, data scientists, machine learning engineers, and developers.

[0164] Machine Learning Engine 1970 can be implemented as a component in Item Listing System 1900, which leverages machine learning algorithms and techniques (e.g., Machine Learning Algorithm 1972) to enhance various aspects of the item listing system's functionality. Machine Learning Engine 1970 can provide a range of machine learning algorithms and techniques for teaching computers to learn from data and make predictions or decisions without explicit programming. These techniques are widely used in a variety of applications across different industries and can include examples such as: supervised learning (e.g., linear regression: classification, support vector machines (SVM); unsupervised learning (e.g., clustering, principal component analysis (PCA), association rules (e.g., apriori); reinforcement learning (e.g., Q-learning, deep Q-networks (DQN); deep learning (e.g., neural networks, convolutional neural networks (CNN), and recurrent neural networks (RNN); and ensemble learning random forests).

[0165] Machine learning training data 120 supports the process of building, training, and fine-tuning machine learning models. Machine learning training data 120 consists of labeled datasets used to teach machine learning models to recognize patterns, make predictions, or perform specific tasks. Training data typically includes two main components: input features (X) and labels or target values ​​(Y). Input features can include variables, attributes, or characteristics that serve as input to the machine learning model. Input features (X) can be numerical, categorical, or even textual, depending on the nature of the problem. For example, in a model used to predict house prices, input features might include the number of bedrooms, square feet, neighborhood, etc. Labels or target values ​​(Y) include the values ​​that the model aims to predict or classify. Labels represent the expected output or basic fact for each corresponding set of input features. For example, in a spam classifier, labels would indicate whether each email is spam (i.e., binary classification). The training process involves presenting training data to the model, and the model learns to make predictions or decisions by recognizing patterns and relationships between the input features (X) and target values ​​(Y). Machine learning algorithms adjust their internal parameters during training to minimize the discrepancy between their predictions and the actual labels in the training data. Machine Learning Engine 1970 can use historical and real-time data to train models and make predictions, thereby continuously improving performance and user experience.

[0166] Machine Learning Engine 1970 can include machine learning models (e.g., Machine Learning Model 1976) generated using the machine learning engine workflow. Machine Learning Model 1976 can include both generative AI models and traditional AI models, both of which can be used in the item listing system 1900. Generative AI models are designed to generate new data (typically in the form of text, images, or other media) based on patterns and knowledge learned from existing data. Generative AI models can be used in a variety of ways, including: content generation, product image generation, personalized product recommendations, natural language chatbots, and content summarization. Traditional AI models encompass a wide range of algorithms and techniques and can be used in a variety of ways, including: recommender systems, predictive analytics, search algorithms, fraud detection, customer segmentation, image classification, natural language processing (NLP), and A / B testing and optimization. In many cases, a combination of both generative and traditional AI models can be used to deliver a comprehensive and efficient e-commerce experience, combining data-driven insights and creativity.

[0167] The Machine Learning Engine 1970 can be used to analyze data, make predictions, and automate processes to provide users with a more personalized and efficient shopping experience. Examples include product recommendation search and filtering; pricing optimization; inventory and stock management; customer segmentation; customer churn prediction and retention; fraud detection; sentiment analysis; customer support and chatbots; image and video analytics; and ad targeting and marketing. The specific applications of machine learning within the Project List Platform 1910 can vary depending on specific objectives, available data, and resources.

[0168] Example Distributed Computing System Environment

[0169] Now refer to Figure 20 , Figure 20 An example distributed computing environment 2000 that can be implemented using the embodiments of this disclosure is shown. Specifically, Figure 20 A high-level architecture of an example cloud computing platform 2010 is shown, which can host a technology solution environment or a portion thereof (e.g., a data trustee environment). It should be understood that this and other arrangements described herein are illustrative only. For example, as mentioned above, many of the elements described herein can be implemented as discrete or distributed components or combined with other components, and implemented in any suitable combination and location. Other arrangements and elements (e.g., machines, interfaces, functions, commands, and function groups) may be used in addition to or in place of the arrangements and elements shown.

[0170] Data centers can support distributed computing environments 2000, including cloud computing platforms 2010, racks 2020, and nodes 2030 (e.g., computing devices, processing units, or blades) within the racks 2020. A technology solution environment can be implemented using a cloud computing platform 2010 that runs cloud services across different data centers and geographic regions. The cloud computing platform 2010 can implement infrastructure controller 2040 components for providing and managing the resource allocation, deployment, upgrades, and management of cloud services. Typically, the cloud computing platform 2010 is used to store data or run service applications in a distributed manner. The cloud computing infrastructure 2010 in the data center can be configured to host and support the operation of endpoints for specific service applications. The cloud computing infrastructure 2010 can be a public cloud, a private cloud, or a dedicated cloud.

[0171] Node 2030 may be configured with host 2050 (e.g., operating system or runtime environment) that runs a defined software stack on node 2030. Node 2030 may also be configured to perform specialized functions (e.g., compute node or storage node) within cloud computing platform 2010. Node 2030 is allocated to run one or more portions of a tenant's service application. A tenant may refer to a customer utilizing the resources of cloud computing platform 2010. The tenant-specific service application components of cloud computing platform 2010 may be referred to as multi-tenant infrastructure or lease. In this document, the terms service application, application, or service are used interchangeably and broadly refer to any software or software portion that runs on top of or accesses storage and compute equipment locations within a data center.

[0172] When node 2030 is supporting more than one individual service application, node 2030 can be partitioned into virtual machines (e.g., virtual machine 2052 and virtual machine 2054). Physical machines can also run individual service applications simultaneously. Virtual machines or physical machines can be configured as personalized computing environments supported by resources 2060 (e.g., hardware and software resources) in the cloud computing platform 2010. It is envisioned that resources can be configured for specific service applications. Furthermore, each service application can be divided into functional parts, allowing each functional part to run on a separate virtual machine. In the cloud computing platform 2010, multiple servers can be used to run service applications and perform data storage operations in a cluster. Specifically, servers can perform data operations independently but are exposed as a single device referred to as a cluster. Each server in the cluster can be implemented as a node.

[0173] Client device 2080 can connect to service applications in the cloud computing platform 2010. Client device 2080 can be configured to correspond to a reference... Figure 20 Any type of computing device described in the computing device 2000, such as client device 2080, can be configured to issue commands to the cloud computing platform 2010. In embodiments, client device 2080 can communicate with service applications via Virtual Internet Protocol (IP) and load balancers or other means that direct communication requests to a specified endpoint in the cloud computing platform 2010. Components of the cloud computing platform 2010 can communicate with each other via a network (not shown), which may include, but is not limited to, one or more local area networks (LANs) and / or wide area networks (WANs).

[0174] Example computing environment

[0175] Having briefly described an overview of embodiments of the present invention, the following describes example operating environments in which embodiments of the present invention can be implemented, in order to provide a general context for various aspects of the present invention. Specifically, first refer to Figure 21An example operating environment for implementing embodiments of the present invention is shown and is generally designated as computing device 2100. Computing device 2100 is merely an example of a suitable computing environment and is not intended to imply any limitation on the scope or functionality of the invention. Nor should computing device 2100 be construed as having any dependency or requirement associated with any one or combination of the components shown.

[0176] This invention can be described in the general context of computer code or machine-usable instructions, including computer-executable instructions (such as program modules) that are executed by a computer or other machine, such as a personal data assistant or other handheld device. Generally, a program module, including routines, programs, objects, components, data structures, etc., refers to code that performs a specific task or implements a specific abstract data type. This invention can be practiced in a variety of system configurations, including handheld devices, consumer electronics, general-purpose computers, and more specialized computing devices. This invention can also be practiced in distributed computing environments, where tasks are performed by remote processing devices linked via a communication network.

[0177] Reference Figure 21 The computing device 2100 includes a bus 2110 that directly or indirectly couples to the following devices: memory 2112, one or more processors 2114, one or more presentation components 2116, input / output ports 2118, input / output components 2120, and a schematic power supply 2122. The bus 2110 represents one or more buses (such as an address bus, a data bus, or a combination thereof). For clarity of concept, Figure 21 The various boxes are shown with lines, and other arrangements of the described components and / or component functions are also envisioned. For example, a presentation component such as a display device can be considered as an I / O component. Additionally, a processor has memory. We recognize this as essential to the art and reiterate... Figure 21 The figures only illustrate example computing devices that can be used in conjunction with one or more embodiments of the present invention. There is no distinction between such categories as “workstation,” “server,” “laptop,” “handheld device,” etc., because all these categories are... Figure 21 Within the scope and with reference to "Computing Devices".

[0178] Computing device 2100 typically includes a variety of computer-readable media. Computer-readable media can be any available media accessible by computing device 2100, and includes volatile and non-volatile media, removable and non-removable media. By way of example, and not limitation, computer-readable media can include computer storage media and communication media.

[0179] Computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to: RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, Digital Universal Optical Disc (DVD) or other optical disc storage devices, magnetic tape cassettes, magnetic tape, disk storage devices or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by computing device 2100. Computer storage media itself does not include signals.

[0180] Communication media typically embody computer-readable instructions, data structures, program modules, or other data in the form of modulated data signals (such as carrier waves or other transmission mechanisms), and include any information transmission medium. The term "modulated data signal" refers to a signal whose characteristics are set or altered in a manner that encodes information in the signal. By way of example, and not limitation, communication media include wired media such as wired networks or direct wired connections, and wireless media such as acoustic, RF, and infrared, as well as other wireless media. Any combination of the above should also be included within the scope of computer-readable media.

[0181] Memory 2112 includes computer storage media in the form of volatile memory and / or non-volatile memory. Memory can be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard disk drives, optical disk drives, etc. Computing device 2100 includes one or more processors that read data from various entities such as memory 2112 or I / O components 2120. Presentation component 2116 presents data indications to a user or other device. Exemplary presentation components include display devices, speakers, printing components, vibration components, etc.

[0182] I / O port 2118 allows computing device 2100 to be logically coupled to other devices, some of which may be built-in, including I / O components 2120. Illustrative components include microphones, joysticks, game controllers, satellite antennas, scanners, printers, wireless devices, etc.

[0183] Additional structural and functional features of embodiments of the technical solution

[0184] Having identified the various components used herein, it should be understood that any number of components and arrangements can be employed to achieve the desired functionality within the scope of this disclosure. For example, for clarity of concept, components in the embodiments depicted in the accompanying drawings are shown in lines. Other arrangements of these and other components can also be implemented. For example, although some components are depicted as single components, many elements described herein can be implemented as discrete or distributed components or combined with other components, and implemented in any suitable combination and location. Some elements can be omitted entirely. Furthermore, as described below, the various functions performed by one or more entities described herein can be performed by hardware, firmware, and / or software. For example, various functions can be performed by a processor executing instructions stored in memory. Therefore, other arrangements and elements (e.g., machines, interfaces, functions, commands, and function groups) can be used in addition to or in place of the arrangements and elements shown.

[0185] The embodiments described in the following paragraphs can be combined with one or more of the specifically described alternatives. Specifically, the claimed embodiments may include references to more than one other embodiment in the alternatives. The claimed embodiments may specify further limitations on the claimed subject matter.

[0186] This document specifically describes the subject matter of embodiments of the invention to meet legal requirements. However, this specification itself is not intended to limit the scope of this patent. Rather, the inventors have envisioned that the claimed subject matter may also be embodied in other ways in combination with other prior art or future art to include different steps or combinations of steps similar to those described in this document. Furthermore, although the terms “step” and / or “box” may be used herein to denote different elements of the method employed, these terms should not be construed as implying any particular order of the various steps disclosed herein, unless and only if the order of the various steps is explicitly described.

[0187] For the purposes of this disclosure, the word "comprising" has the same broad meaning as the word "including," and the word "access" includes "receiving," "referencing," or "retrieval." Furthermore, the word "communication" has the same broad meaning as the words "receiving" or "transmitting," which is facilitated by a software or hardware-based bus, receiver, or transmitter using the communication medium described herein. Additionally, unless otherwise indicated, words such as "a" or "an" include both plural and singular forms. Thus, for example, the constraint of "feature" is satisfied when one or more features are present. Furthermore, the term "or" includes conjunctions, disjuncts, and both (therefore, a or b includes a or b as well as a and b).

[0188] For the purposes of the detailed discussion above, embodiments of the present invention are described with reference to a distributed computing environment; however, the distributed computing environment described herein is merely exemplary. Components may be configured to perform novel aspects of the embodiments, wherein the term "configured for" may mean "programmed to" perform a specific task or implement a specific abstract data type using code. Furthermore, while embodiments of the present invention can generally be referred to the technical solution environment and schematic diagrams described herein, it should be understood that the described techniques can be extended to other implementation contexts.

[0189] Embodiments of the invention have been described with respect to specific embodiments intended to be illustrative and not limiting in all respects. Alternative embodiments will become apparent to those skilled in the art without departing from the scope of the invention.

[0190] As can be seen from the above, the present invention is well suited to achieving all the objectives and purposes set forth herein, as well as other obvious and inherent advantages of the structure.

[0191] It should be understood that certain features and sub-combinations are practical and can be adopted without reference to other features or sub-combinations. This is contemplated by the claims and is within the scope of the claims.

Claims

1. A computerized system, comprising: One or more computer processors; as well as A computer memory that stores computer-usable instructions, which, when used by the one or more computer processors, cause the one or more computer processors to perform operations, the operations including: Access the project list associated with the project list system; The list of projects was analyzed using a brand-oriented machine learning model trained on a multidimensional authenticity analysis dataset and brand multidimensional authenticity features; Based on the analysis of the project list, a brand-oriented safety notification associated with the project list is generated; and The brand-oriented security notification is transmitted.

2. The system according to claim 1, wherein, The brand-oriented machine learning model is integrated into the brand-oriented list management tool of the project list system to support brand-oriented list management.

3. The system according to claim 1, wherein, The multidimensional authenticity analysis dataset includes intellectual property enforcement data, certified souvenir data, and item listing system data, which provide basic fact labels and annotations associated with model training.

4. The system according to claim 1, wherein, The brand-oriented machine learning model is trained based on machine learning technology, which integrates multi-dimensional brand authenticity features corresponding to the intellectual property enforcement data, the certified souvenir data, and the project listing system data.

5. The system according to claim 1, wherein, The brand-oriented security notification is generated based on the predicted brands identified in the list of items, wherein the list of items does not include brand input.

6. The system according to claim 1, wherein the operation further includes: Based on the brand-oriented security notification, a list quality analysis is performed, which is in part based on brand-oriented security data associated with the brand-oriented security notification; as well as Based on the quality analysis of the list, the list of items is flagged for remedial action.

7. The system according to claim 1, wherein, The brand-oriented security notification is sent to the sellers associated with the item list and the security administrator of the item list system.

8. One or more computer storage media having computer-executable instructions thereon, said computer-executable instructions, when executed by a computing system having a processor and a memory, causing the processor to perform operations, said operations including: Access the project list associated with the project list system; The list of projects was analyzed using a brand-oriented machine learning model trained on a multidimensional authenticity analysis dataset and brand multidimensional authenticity features; Based on the analysis of the project list, a brand-oriented safety notification associated with the project list is generated; and The brand-oriented security notification is transmitted.

9. The medium according to claim 8, wherein, The brand-oriented machine learning model is integrated into the brand-oriented list management tool of the project list system to support brand-oriented list management.

10. The medium according to claim 8, wherein, The multidimensional authenticity analysis dataset includes intellectual property enforcement data, certified souvenir data, and item listing system data, which provide basic fact labels and annotations associated with model training.

11. The medium according to claim 8, wherein, The brand-oriented machine learning model is trained based on machine learning technology, which integrates multi-dimensional brand authenticity features corresponding to the intellectual property enforcement data, the certified souvenir data, and the project listing system data.

12. The medium according to claim 8, wherein, The brand-oriented security notification is generated based on the predicted brands identified in the list of items, wherein the list of items does not include brand input.

13. The medium according to claim 8, wherein the operation further comprises: Based on the brand-oriented security notification, a list quality analysis is performed, which is in part based on brand-oriented security data associated with the brand-oriented security notification; as well as Based on the quality analysis of the list, the list of items is flagged for remedial action.

14. The medium according to claim 8, wherein, The brand-oriented security notification is sent to the sellers associated with the item list and the security administrator of the item list system.

15. A computer-implemented method, the method comprising: Access the project list associated with the project list system; The list of projects was analyzed using a brand-oriented machine learning model trained on a multidimensional authenticity analysis dataset and brand multidimensional authenticity features; Based on the analysis of the project list, a brand-oriented safety notification associated with the project list is generated; and The brand-oriented security notification is transmitted.

16. The method according to claim 15, wherein, The multidimensional authenticity analysis dataset includes intellectual property enforcement data, certified souvenir data, and item listing system data, which provide basic fact labels and annotations associated with model training.

17. The method according to claim 15, wherein, The brand-oriented machine learning model is trained based on machine learning technology, which integrates multi-dimensional brand authenticity features corresponding to the intellectual property enforcement data, the certified souvenir data, and the project listing system data.

18. The method according to claim 15, wherein, The brand-oriented security notification is generated based on the predicted brands identified in the list of items, wherein the list of items does not include brand input.

19. The method of claim 15, further comprising: Based on the brand-oriented security notification, a list quality analysis is performed, which is in part based on brand-oriented security data associated with the brand-oriented security notification; as well as Based on the quality analysis of the list, the list of items is flagged for remedial action.

20. The method of claim 15, wherein, The brand-oriented security notification is sent to the sellers associated with the item list and the security administrator of the item list system.

Images